Ueland), Centre for International Health (Dr Kv\l=a%o\le),. University of Bergen ... Norway (Dr Nyg\l=a%o\rd). ...... D'Avanzo B, Santoro L, La Vecchia C, et al.
Total Plasma Homocysteine and Cardiovascular Risk Profile The Hordaland Ottar
Stein Emil Vollset, MD, DrPH; Helga Refsum, MD; Inger Stensvold, MSc; Nordrehaug, MD; Per Magne Ueland, MD; Gunnar Kv\l=a%o\le,MD
Objective.\p=m-\Toestimate the relations between established cardiovascular risk homocysteine (tHcy) in plasma. Design.\p=m-\Healthexamination survey by the Norwegian Health Screening Ser-
factors and total
vice in 1992 and 1993.
Setting.\p=m-\Generalcommunity, Hordaland County of Western Norway. Participants.\p=m-\Atotal of 7591 men and 8585 women, 40 to 67 years of age, with no history of hypertension, diabetes, coronary heart disease, or cerebrovascular
disease were included. Main Outcome Measure.\p=m-\PlasmatHcy level. Results.\p=m-\Thelevel of plasma tHcy was higher in men than in women and increased with age. In subjects 40 to 42 years old, geometric means were 10.8 \g=m\mol/L for 5918 men and 9.1 \g=m\mol/L for 6348 women. At age 65 to 67 years, the corresponding tHcy values were 12.3 \g=m\mol/L(1386 men) and 11.0 \g=m\mol/L(1932 women). Plasma tHcy level increased markedly with the daily number of cigarettes smoked in all age groups. Its relation to smoking was particularly strong in women. The combined effect of age, sex, and smoking was striking. Heavy-smoking men aged 65 to 67 years had a mean tHcy level 4.8 \g=m\mol/L higher than never-smoking women aged 40 to 42 years. Plasma tHcy level also was positively related to total cholesterol level, blood pressure, and heart rate and inversely related to physical activity. The relations were not substantially changed by multivariate adjustment, including intake of vitamin supplements, fruits, and vegetables. Conclusions.\p=m-\Elevatedplasma tHcy level was associated with major components of the cardiovascular risk profile, ie, male sex, old age, smoking, high blood pressure, elevated cholesterol level, and lack of exercise. These findings should influence future studies on the etiology and pathogenesis of cardiovascular disease. (JAMA. 1995;274:1526-1533)
PATIENTS with homocystinuria, a rare inborn error of metabolism, have ex¬ tremely high levels of total homocyste¬ ine (tHcy) in plasma. They also have a
From the Section for Medical Informatics and StatisNyg\l=a%o\rdand Vollset), Department of Clinical Biology, Division of Pharmacology (Drs Refsum and Ueland), Centre for International Health (Dr Kv\l=a%o\le), University of Bergen, and Department of Heart Disease, Haukeland University Hospital, Bergen, Norway (Dr Nordrehaug), and the National Health Screening Service, Oslo, Norway (Dr Tverdal and Ms Stensvold). Reprint requests to Section for Medical Informatics and Statistics, Armauer Hansens hus, 5021 Bergen,
Norway (Dr Nyg\l=a%o\rd).
high incidence of cardiovascular disease in early adolescence and even in child¬ hood. Homocystinuria results from sev¬ eral enzymic defects in homocysteine (Hey) metabolism, but premature car¬ diovascular disease develops irrespec¬ tive of the site of metabolic deletion,1 suggesting that some form of Hey is responsible for the vascular damage.2 Plasma tHcy refers to the sum of pro¬ tein-bound, free-oxidized, and reduced species of Hey in plasma3 and is usually about 5 to 15 µ /L in healthy sub¬ jects.4 Moderate hyperhomocysteinemia (15 to 30 µ /L)5 is related to genetic
Downloaded From: http://jama.jamanetwork.com/ by a University of Oslo User on 06/19/2013
Aage Tverdal, PhD;
acquired factors.4 Genetic causes are heterozygous deficiency of cystathionine ß-synthase or methylenetetrahydrofoor
late reducíase or a thermolabile variant of the latter enzyme,5 whereas impaired cobalamin or folate status is among the
Since the first report in 1976,8 more than 20 retrospective5·911 and two pro¬ spective12·13 studies of more than 3300 patients have demonstrated a relation between moderate hyperhomocysteinemia and premature vascular disease in the coronary, cerebral, and peripheral arteries. Most studies conclude that plasma Hey is an independent risk fac¬ tor for cardiovascular disease,5·9·1416 but an association between Hey levels and established cardiovascular risk factors, such as serum cholesterol level,17·18 blood pressure,18·19 or cigarette smoking,13·20 has occasionally been demonstrated. Knowl¬ edge of such associations is needed to identify potential confounders in stud¬ ies of Hey and disease and may contrib¬ ute toward understanding of the pathogenesis of cardiovascular disease. Recently, we initiated a population-
based, prospective study on plasma tHcy
and cardiovascular disease. In the cur¬ rent report, we have used the baseline data to estimate the relations between plasma tHcy level and major components of the cardiovascular risk profile. Our study is the first to address these issues in a large number of subjects.
SUBJECTS AND METHODS
Study Population The study, which is part of a national
cardiovascular risk survey,21 was con¬ ducted in the Hordaland County in West¬ ern Norway from April 1992 to April 1993 by the National Health Screening
1.—Population Size, Attendance Rates, and Invited No.
Characteristic Sex Male
Numbers Included in the Attendance Rate, No. (%) 8573
12 594(72.8) 683
(68.4) 4766 (73.2) 18043(72.7)
Study Study Subjects, No.
12 266(70.9) 592 (59.3) 3318(50.9) 16176(65.2)
"Includes all who attended without reporting a history of diabetes mellitus, angina, former acute myocardial infarction, stroke, or treatment for hypertension. One subject with homocystinuria was excluded.
Service (NHSS) in cooperation with the University of Bergen and local health services. The total population of Hordaland is approximately 420 000, of whom about 50% live in the city of Bergen. The eligible subjects were selected from the National Population Registry, identified by site of residence and their age on December 31, 1992. A total of 24 815 subjects from three different age groups were invited to participate in the study. The younger age group in¬ cluded all subjects in the county who were 40 to 42 years of age. The older age group covered all subjects aged 65 to 67 years in Bergen and three neighboring suburban municipalities. A third group (43 to 64 years) was a 2% random sample of residents in Bergen. The attendance rate for the whole group was 72.7%. All participating sub¬ jects gave their written, informed con¬ sent, and the study protocol was ap¬ proved by the Regional Ethical Com¬ mittee of Western Norway. A total of 1866 participants reporting a previous diagnosis of coronary heart disease, cerebrovascular disease, hyper¬ tension, or diabetes were excluded from the analyses to avoid potential influence of the disease itself on tHcy levels di¬ rectly or indirectly from treatment or from change in lifestyle. One patient with homocystinuria was also excluded. Thus, 16176 individuals were included in the current
study (Table 1).
Data Collection The screening procedures performed by the NHSS have been described pre¬
through questionnaires, examinations, and blood tests. The subjects received
the invitation letter and a one-page ques¬ tionnaire. This was completed at home by the participants and collected and
checked for logical errors by a nurse on the day of examination. The question¬ naire provided information about type of work, physical activity, smoking hab¬ its, medical history of cardiovascular dis¬ ease, hypertension and diabetes mellitus, and family history of coronary heart disease. A second questionnaire about
recent food intake was filled in by a nurse during the examination. The at¬
tending subjects also received a third questionnaire on the examination day, covering more details about lifestyle, medical history, and dietary habits. This was later completed at home and mailed to the NHSS by 86% of the participants.
Smoking Habits.—Current and former cigarette smokers were asked to report the average number of cigarettes smoked per day and the duration of smoking. Based on this information, ciga¬ rette smokers were grouped in five cat¬ egories, ie, never smokers, former smok¬ ers, light smokers (one to nine cigarettes per day), moderate smokers (10 to 19 cigarettes per day), and heavy smokers (>20 cigarettes per day).
Physical Activity.—The subjects asked to mark one of the following categories that best fitted their average degree of activity in leisure time for the last year: (1) sedentary or no activity; (2) walking, cycling, or other type of moderate physical activity for at least 4 hours a week (moderate activity); (3) exercise, gardening with physical exer¬ tion, or similar degree of physical ac¬ tivity for at least 4 hours a week (active exercise); or (4) regular heavy training or competitive sport several times a week (heavy training). The grading of leisure time physical activity obtained by the questionnaire was validated by the finding of a strong inverse relation between triglycérides and reported activity level. Low level of triglycérides is a recognized response to increased physical activity in leisure
Intake of Vitamin Supplements, Fruits, and Vegetables.—A vitamin supplement score (n=11940) was cre¬
ated according to use during the year and frequency of intake during the week and categorized into five groups. The lowest category included subjects who never used vitamins (34%), while the highest category represented those who took vitamins 6 to 7 days a week during the whole year (14%). Among those tak¬ ing vitamins, approximately 80% re¬ ported use of vitamin and/or multivi-
Downloaded From: http://jama.jamanetwork.com/ by a University of Oslo User on 06/19/2013
tamin combinations, which usually con¬ tain folie acid (100 µg per tablet). A fruit-vegetable score (n=13 378) was based on the sum of frequencies of in¬ take of fruits and vegetables (four cat¬ egories). The highest category repre¬ sented subjects consuming both fruits and vegetables at least 6 days a week (38%), while the lowest category rep¬ resented those with intake of fruits and vegetables once a week or less (4.3%). In a subpopulation (n=329), we found a significant correlation between plasma fo¬ late and the vitamin supplement score (r=0.31) and the fruit-vegetable score (r=0.12). No relation was observed be¬ tween these scores and plasma cobalamin. Another score based on intake of vegetables and fruits, which are rich in folates, showed a stronger relation to plasma folate (r=0.21) but a weaker re¬ lation to plasma tHcy (r=0.10) than the fruit-vegetable score (r=—0.14), and it was therefore not included in the analyses. Examinations Trained nurses and technical staff from the NHSS performed the exami¬ nations and collected the blood samples. After the participants had registered and height and weight were measured, they were allowed to sit for 10 minutes while the nurse checked the first ques¬ tionnaire and obtained information for the second questionnaire. Then, three blood pressure measurements using Dinamap 845 XT equipment (Criticón, Tampa, Fla) were performed. Values from the second measurement were used in this study. Blood Sample Collection and Biochemical Analysis
The participants were nonfasting, and blood samples were obtained immedi¬ ately after the examination with the per¬ son in the sitting position. Blood samples used for the preparation of serum were collected into an evacuated tube con¬ taining sodium sulfite titration gel (Beeton Dickinson Co, Meylan, France) and centrifuged within 2 hours. The serum tubes were transported to the Depart¬ ment of Clinical Chemistry, Ullevâl Hos¬ pital, Oslo, where determination of total cholesterol and triglycérides was per¬ formed within 7 days after collection of the sample. Blood samples used for the prepara¬ tion of plasma were collected into an evacuated tube containing ethylenediaminetetraacetic acid, placed in a re¬ frigerator (4°C to 5°C) within 15 to 30 minutes, and centrifuged usually within 1 hour (maximum, within 3 hours). The plasma fraction was then transferred to plastic vials. Compared with immediate sample handling, the described proce-
Table 2.—Total Plasma
tHcy, µ No.
tHcy, µ No.
9.14(9.06-9.22) 9.60 (9.50-9.69)
Geometric mean Arithmetic mean
*CI indicates confidence interval.
plasma tHcy level of 0.5 µ /L.25 The unfrozen plasma samples, kept at room temperature, usually arrived at our labo¬ ratory the following day, but occasion¬ ally they arrived up to 4 days after col¬ lection. The samples were stored at -20°C until the tHcy analysis was per¬ formed. The duration of storage was be¬ tween a few days and up to 6 months. Total plasma Hey, cysteine, and cysteinylglycine levels were determined us¬ ing a modification25 of a fully automated assay based on precolumn derivatization with monobromobimane followed
by reversed-phase high-performance liq¬ uid chromatography.25·26 The precision (between-day coefficient of variation) of
the assay is less than 3%, and results obtained correlate well with other es¬ tablished methods for determination of tHcy level.4 The stability of analytical procedures was controlled for by insert¬ ing quality-control samples for every 18th sample. Replicate measurements were performed for about every 15th sample. This included all samples with very high (&40.0 µ /L) or low (